analysis
play

Analysis Manuel Len Hoyos Overview What is Time Series Data? - PowerPoint PPT Presentation

Time Series Analysis Manuel Len Hoyos Overview What is Time Series Data? Index Prices: Crude Oil Gold Bitcoin What is Time Series Analysis? Uses Forecasting Time Series Data A collection of observations of a


  1. Time Series Analysis Manuel León Hoyos

  2. Overview ❖ What is Time Series Data? Index Prices: ➢ Crude Oil ➢ Gold ➢ Bitcoin ❖ What is Time Series Analysis? ➢ Uses ➢ Forecasting

  3. Time Series Data ➢ A collection of observations of a particular variable made chronologically. - Numerical - Same time intervals - Large in size ➢ Examples: Webster University enrollment per year, Gross Domestic Product (GDP), population census, unemployment rate, daily temperature, etc.

  4. Time Series Analysis ▪ Methods for analyzing time series data in order to extract meaningful statistics and other characteristics of the data. ➢ Interpretation ➢ Forecasting ➢ Hypothesis testing ➢ Trend analysis ➢ Control (response) ➢ Simulations Fields: economics, finance, geology, meteorology, business, biology, etc.

  5. Uses of Time Series Analysis ▪ Description (monitoring data) -Describe patterns over time ▪ Explanation -Consider all possible factors in understanding the behavior of a series ▪ Forecasting -Prediction of future values based on the past - Helpful for business decisions: production, inventory, personal, etc. ▪ Improving past behavior -Identifying factors influencing. Example: action over increasing levels of air pollution

  6. Trend Analysis ▪ Sustained movements in the variable of interest in a specific direction. ▪ Horizontal pattern (mean) ▪ Trend pattern (upwards or downwards) ▪ Season pattern (depending on weather or frequency of events) ▪ Cyclical pattern (Up, down, up, …)

  7. Oil Prices (per barrel) Historical max: $145 July, 2008

  8. Volatility of Oil Prices

  9. Forecasting ➢ Estimating how a series of observations will continue in the future ➢ Considering current and past values ➢ Models assume the future will show patterns from the past ✓ Uncertainty about the future ✓ Easier to forecast in the short-term

  10. ARMA & ARIMA Models (Hyndman, 2017. Forecasting in R )

  11. Gold Prices (per ounce) Historical Max: $1,895 September, 2011

  12. Forecasting Gold Prices

  13. Bitcoin Prices Historical Max: $19,187 December 16, 2017

  14. Forecast of Bitcoin Expected to cross $25,000 in 12 days

  15. Summary ❖ What is Time Series Data? ❖ What is Time Series Analysis? ➢ Uses ➢ Forecasting

  16. References Bennett, R. & Hugen, D. (2016). Financial Analytics with R. Cambridge University Press. Brockwell-Davis (2016). Introduction to Time Series and Forecasting . Springer. Cowpertwait & Metcalfe (2009). Introductory Time Series with R. Springer. Hyndman, R. (2017). Forecasting in R . Data Camp. Singh, A. & Allen, D (2017). R in Finance and Economics A Beginner’s Guide . World Scientific. Wikipedia. (2017). Autoregressive integrated moving average. https://en.wikipedia.org/wiki/Autoregressive_integrated_moving_average Wikipedia (2017). Time Series. https://en.wikipedia.org/wiki/Time_series Wikipedia (2017). Stochastic Process. https://en.wikipedia.org/wiki/Stochastic_process

Recommend


More recommend